Blogg

• July 31, 2026

  A URL that wakes your agent

  Agents wake when you message them or when their calendar says it is time. Now anything that can send an HTTP POST can wake them too: an n8n workflow, a Zap, a GitHub Action.

  Läs → 4 min

• Building a useful agent · Memory · July 29, 2026

  Understanding an AI agent's memory through Harry Potter

  Compare what you remember from the Philosopher's Stone to what our agent remembers.

  Läs → 3 min

• Building a useful agent · Memory · July 28, 2026

  How memories become knowledge

  You go to bed with a day of moments and wake up knowing things. Our agents do the same: a nightly pass turns repeated moments into knowledge that lasts.

  Läs → 7 min

• Building a useful agent · Memory · July 21, 2026

  How a conversation becomes a memory

  Some experiences stay with you for life. Others disappear almost as soon as they happen. We built our agents' memory to work the same way.

  Läs → 6 min

• Building a useful agent · Memory · July 9, 2026

  Memory and dreams: creating the next generation of AI assistants

  Getting access to ChatGPT felt like having an expert you could ask anything. But it was a new expert every time. You always started from the beginning. A good colleague or friend is rarely the one who knows the most; they are the one who knows you, your work, your situation. What they have that the expert doesn't is a shared history. Memories.

  Läs → 6 min

• July 2, 2026

  Mistral vs. Claude on our onboarding: 4× faster, 30% cheaper

  We put Mistral Medium 3.5 up against Claude Sonnet 4.6, the model behind our onboarding agent, on our own workload. Mistral came out about 4× faster and 30% cheaper, at near-equal quality. For this particular job it turned out to be a surprisingly good fit, for reasons specific to how onboarding actually runs.

  Läs → 4 min